Catalyst-Free Crosslinking Modication of Nata-De-Coco-Based Bacterial Cellulose Nanobers Using Citric Acid

Bacterial cellulose (BC) has gained research attention in materials science and biomedicine due to its fascinating properties. BCs' ber collapse phenomenon (inability to reabsorb water after dehydration) is one of the drawbacks that limit its potentials. To overcome this, a catalyst-free thermal crosslinking reaction was employed to modify the BC using citric acid (CA) without compromising the biocompatibility. Properties evaluation of the modied BC (MBC) by FTIR, XRD, SEM/EDX, TGA, and Tensile analysis conrmed the ber crosslinking and improvement of some properties that could be advantageous for various applications. The modied nanober seems to maintain its inherent crystallinity and thermal stability with an increased water absorption/swelling and tensile modulus. The resulting MBC reported here can be relevant for wound dressings and tissue scaffolding.


Introduction
The Philippines originated jelly dessert (Nata-de-coco) is the cheapest form of bacterial cellulose (BC) produced through the fermentation of coconut water [1]. It is a pure form of (BC) with unique physicochemical, morphological, and mechanical properties [1,2]. Owing to this, nata-de-coco-based BC can serve as a good reference point for applications such as biomedicine, where high material purity is a primary demand. Its large-scale production in many Asian countries like Malaysia, Indonesia, and Thailand [3] will not be unconnected with the advantageous ease of the process.
BC is a micro brillar biomaterial rst reported by Andrian Brown in 1886. Different species of bacteria such as Acetobacter xylinum, Rhizobium, Achromobacter, and Sarcina were reported to produce BC through fermentation [4][5][6]. BC is generally pure, biocompatible, non-toxic, and has a broad modifying ability into different forms and compositions. This earns it a remarkable range of applications in many elds of science and medicine [7][8][9]. However, its inability to reabsorb much water after being dehydrated and inadequate functionality limits the said attractiveness, hence the need for its modi cation [10].
In-situ and ex-situ modi cations were principally the major approaches known for BC modi cation. The ex-situ modi cation mostly after BCs' production, involves either physical or chemical process [11][12][13].
One of the methods of interest is the crosslinking reaction (a process that induces a strong linkage between the polymer chains through covalent bonding), owing to its simplicity and effectiveness [14].
CA is one of the organic acids enlisted as generally regarded as safe (GRAS) by the US food and drug administration (FDA) that has long been used as modi er on many polymeric biomaterials via crosslinking, including the BC [15][16][17][18][19]. Modi cation of polymeric biomaterial with CA mostly yielded what is also known as citrate-based biomaterial (CBB) [20,21]. Owing to their excellent biological, chemical, and material properties, such as antimicrobial, antioxidant, and uorescence properties, CBBs have been trending in biomedicine [22][23][24]. Crosslinking a biopolymer with CA requires an elevated temperature of 120-190 ºC [25]. This method offers the advantage of ne-tuning the material properties such as mechanical, chemical, and degradation properties [26,27]. CA crosslinking of biopolymers involving different catalysts has been reported by many scientist, but the undesirable effects posed by the catalysts [28, 29] may limit their application in biomedicine. We, therefore, hypothesised that "the use of the catalyst for BCs' crosslinking modi cation is unnecessary". In this study, a catalyst-free thermal crosslinking approach was employed for the rst time to modify the nata-de-coco-based BC bers using a readily available and inexpensive multifunctional monomer, the CA The physicochemical, morphological, and mechanical properties evaluation of the resulting biopolymer showed that it retained most of its important properties such as crystallinity and thermal stability and improved water absorption rate and tensile modulus compared to the unmodi ed. Our method seems to be the cheapest and easiest approach that yields promising improvements of BCs' properties and showed that the use of the catalyst for BCs' modi cation might be unnecessary.

Materials And Methods
Bacterial cellulose (BC) sheets were purchased from a local Nata-de-Coco company (Happy Alliance). Citric acid (CA) monohydrate powder (C 6 H 8 O 7 .H 2 O) and Sodium hydroxide (NaOH) were all purchased from Merck (Sigma-Aldrich).

Puri cation and Modi cation of BC
BC hydrogels were modi ed with CA by thermal crosslinking as in [30] with slight modi cation to exclude the catalyst. Brie y, the wet BC sample was cut into 100 mm x 50 mm and puri ed in 1% (w/v) NaOH at 90 ºC for 60 mins then washed with distilled water until the pH becomes 7-8 at 37 ºC. BC sheets were

Scanning Electron Microscopy (SEM.)
The surface morphology of the bers before and after modi cation was examined by SEM analysis (Model: Hitachi TM3000, Japan) equipped with an (EDX) system. Micrographs of platinum sputter-coated were taken at an accelerating voltage of 15 kV for different magni cations. Full-scale elemental quanti cation data were also acquired from the EDX system.

Fourier Transformed Infrared (FTIR)
The BC, MBC samples were analysed using (Model: PerkinElmer-Frontier™, L1280044, USA) spectrophotometer equipped with an attenuated total re ection (ATR-FTIR) system as in [31]. The spectra were obtained from scans between a wavelength range of 4000 to 650 cm − 1 and 4 cm − 1 resolution.
Where I 200 represents the maximum intensity of the diffraction peak at (200) lattice plane and I am represents the diffraction intensity at 2θ = 18 º corresponding to the amorphous region of BC.
K is the Scherrer's constant (0.9), λ is the X-ray wavelength (1.54Å), FWHM is the width of the diffraction peak at maximum height, and θ is the Bragg's angle.
The term Z denotes discriminant function developed by [33], d 1 is the d-spacing of 1-10 peak, and d 2 is the d-spacing of 110 peaks. Where Z > 0 or Z < 0 indicates I α or I β rich type of cellulose, respectively [34].

Water Contact Angle (WAC.)
Water contact angle (WCA.) Optima machine (Model: 1020046094) equipped with a camera was used to capture the water droplet image and measure the contact angle. Brie y, 20 x 20 mm sheets were cut after freeze-drying before the process. A uniform droplet of 2.0 µl of deionised water (diH 2 O) was dispensed on ve (5) different points on each sample, and an average angle was then recorded [35].

Swelling rate (SR.)
Freeze-dried samples were cut into 30 x 30 mm, the dried weight (W 1 ) recorded, then immersed in either distilled water or SBF at ambient temperature. Samples were then removed and weighed at certain intervals after been blotted with a lter paper to remove the excess water until an equilibrium weight (W 2 ) is reached. The swelling rate was calculated using Eq. 4 below [36, 37].
Where W 1 is the dried weight and W 2 is the nal wetted weight.

Thermal Gravimetric Analysis (TGA)
The thermal stability of all samples was evaluated using a thermal analyser (Shimadzu DTG-60H, Japan). A freeze-dried lm weighing 18 mg ± 3 mg for all the samples in a platinum pan was heated between 30 ºC and 900 ºC at 10 ºC/min heating rate under a nitrogen atmosphere of 100 ml/min ow rate. The weight loss upon heating was normalised as percentage weight loss (%) and plotted against the corresponding temperature (ºC) [38].

Tensile properties
Tensile properties of the BC and MBC samples were evaluated using a tensile testing machine (Zwick/Roell Z020, Zwick, Germany) according to ASTM-D882 standards. Brie y, freeze-dried samples kept in a desiccator were cut into a rectangular shape (70 x 20 mm) with different thickness. A "gauze" length of 50 mm was used, and all measurements were performed for at least ve samples at a crosshead speed of 2 mm/min [39,40] in an ambient condition. a bit compacted ber similar to the untreated sample. This could be due to the high crosslinking density that occurs between the bers, thus affecting the porosity and preventing the passage of water molecules leading to a low swelling rate, as explained in Section 3.5 and Figure 7.

Results And Discussion
The EDX spectra in Figure 2 represents the elemental composition (Carbon and Oxygen) typical of organic ber, while the elemental and atomic weight percentages were presented in Table 1 for the pure and modi ed BC.
The slight shift of the elemental and atomic weight percentage (Table 1)  increasing that of oxygen, similar to the trend reported for ascorbic acid modi cation on cellulose by [42].
The EDX data showing the carbon (C) and oxygen (O) peaks only (Figure 2) is also suggesting the notion that nata-de-coco-based BC is highly (~99%) pure [43].

Fourier Transformed Infrared (FTIR)
The FTIR spectra of the pristine and modi ed samples were shown in Figure 3. The signature peaks attributed to the dominant functional group of BCs' (OH-stretching) vibration were at 3346 cm -1 . Peaks obtained at 2865 cm -1 and 1420 cm -1 were due to C-H stretching and 1450 cm -1 due to CH 2 absorptions.
Peaks obtained at 1719 cm -1 related to carbonyl/carboxyl (C=O) stretching [44] appears only on the crosslinked samples thus, con rming the presence of CA within the modi ed BC samples [30,45,46].
Peaks between 1055 cm -1 , 1020 cm -1 were due to C-O-C interactions. The reduced intensity of the OH peaks on the crosslinked samples can also result from the chemical interaction with CA [47]. Overall, the low intense OH peaks at 3346 cm -1 and 1711 cm -1 on treated samples are indications that crosslinking modi cation on the BC was successful. The proposed mechanism of CA crosslinking on BC was presented in a schematic diagram in Figure 4.

X-ray Diffraction (XRD)
XRD patterns shown in Figure 5 represents the spectra obtained for the pure, modi ed samples. All The peaks associated with BCs' crystallinity appear with similar intensities for all the samples, indicating that the CA modi cation has less effect on the crystalline structure and morphology of the BC [51,52].
Even though [15] reported a decrease in crystalline peaks on sodium carboxymethylcellulose (NaCMC) crosslinked with CA, such may likely be due to one of the cellulose derivatives. De Lima et al. suggested that the decreased crystallinity they observed is ascribed to the viscosity increase of NaCMC or its interaction with cellulose nano bers during crystallisation [53].
The crystallinity index and crystallite size values calculated from the XRD data were between 92% to 95% and 51Å to 56Å. This essentially shows that the CA crosslinking has less effect on the crystallinity and crystallite size of the MBC. Furthermore, cellulose I allomorph calculated using the Z-discriminant function also showed that all the samples have the same cellulose Iα rich (triclinic) form, the typical of bacterial cellulose [54,55]. All calculated agreed with previously reported data [30] and indicate that the BC still maintained its crystalline nature after the crosslinking modi cation.

Page 7/21
The wetting behaviour of a materials' surface is closely related to the molecular terminal groups present, and contact angle studies give information on the wettability properties of a material [56]. In theory, a surface is considered hydrophilic or super hydrophilic when its WCA is smaller than 90º or 10º, respectively [35]. Figure 6 represents the mean contact angle measured for the pure and modi ed BC samples. All, including the pure BC, fall between 0º and 33.90º, signifying that all samples were either hydrophilic or super hydrophilic depending on their contact angles. However, it noteworthy that the modi ed samples have shown a decreasing WCA up to 0º (MBC0.30 and MBC0.60) where the water droplet is no longer capturable (disperses as soon as dropped). BC's hydrophilicity could be attributed to the additional carboxyl groups [44,57] that can form hydrogen bonds with water molecules [58]. Even though a native BC is inherently hydrophilic, the angle of contact with water tends to decrease with increasing the CA concentration. Essentially here, the CA modi cation on BC has improved its surface chemistry to attract more water further.

Swelling rate (SR.)
Generally, polymeric materials' water absorption and swelling behaviour are through capillary action and diffusion and electrostatic repulsion between the ions on the polymer chains that forces it to expand and swell [57]. The swelling rate (SR) for the pure and modi ed BC samples was presented in Figure 7.
Modi ed samples have shown an increased SR, mostly at lower and decrease higher concentrations of CA to a rate even below that of the pure BC. The decrease in SR with the increase in CA concentration could be due to the numerous crosslinker points formed within the BCs' ber networks, thus reducing the spaces for water to enter [57]. It is evident for samples (MBC0.30 and MBC0.60) having the lowest absorption rates, a packed ber geometry on the SEM micrographs in Figure 1, and the sample images in gure 7, which could result from the high concentration of CA Water absorption/swelling is especially advantageous for BCs' medical applications, such as wound dressings [28]. Interestingly, all samples showed similar SR in both SBF and DI water. The SR results reported here agreed with the previous report that BC water holding capacity is between 60 to 700 times its dry weight [59].

Thermal Gravimetric Analysis (TGA)
An important property of BC is its thermal stability, especially for applications in biomedicine where higher temperatures are applied for sterilisation processes. Figure 8 shows the thermal behavior of the pristine and modi ed BC evaluated in this study. An initial weight loss observed for all the samples at a temperature between 45 to 120 ºC was due to absorbed moisture evaporation. Except for samples with the highest CA concentration (MBC0.30 and MBC0.60) that displayed a partial decomposition between 120 to 300 ºC, all other modi ed samples were not different from the pristine BC. They all showed a maximum weight loss at a temperature between 300 to 392 ºC due to dehydration, decomposition, and dissociation of the glycosidic linkages [60-62]. The partial, total and residual mass loss observed at maximum temperatures 300 ºC, 392 ºC, and 620.93 ºC were 25.928%, 88.149%, and 7.875%, respectively. The partial decomposition observed may also be due to the high concentration of CA that attracts more moisture than the lower concentrations. Our result implies that the CA modi cation has less effect on the thermal properties of the BC [63]. Table 2 represents the tensile test results for all the samples with the modi ed showing improved mechanical strength compared to the unmodi ed, except for the lowest CA concentration (MBC0.03), which exhibits a very low tensile strength value. The decrease in the mechanical strength displayed could result from lesser crosslinking degrees within the ber networks due to the low amount of the crosslinker, which can also be seen from the SEM micrographs Figure 1. The sample thickness and spongy appearance after freeze-drying could also lead to the loosening of the bers. It can also be observed that, despite having a lower modulus value, the elongation at break is within the same range as other modi ed samples, implying that the elasticity of the bers is close to other modi ed samples after reaching the maximum yield limit.Like the modulus, the tensile strength also follows the trend as increasing with the CA concentration if not for the lowest concentration. However, the elongation at break showed a different pattern where it increases from the lowest and decreases at the highest concentration, especially in the lowest concentration. Therefore, it can be hypothesised that high concentration CA treatment on BC may have a negative effect on the stretching ability of the BC ber.

Conclusion
Several attempts to enhance BCs' properties through crosslinking modi cation involve the use of catalysts. Some of these catalysts might alter the chemical composition and compromise the resulting polymer's biocompatibility, limiting its application, especially in biomedicine. Here we reported for the rst time a catalyst-free modi cation of BC with CA through a simple immersion hydrothermal crosslinking. The improvements in the chemical, morphological, thermal, and mechanical properties presented in this report are an indication that the modi cation has resulted in a potential citrate-based biopolymer that can be used as future wound dressings or a tissue scaffold material. The approach used here seems to be the cheapest and easiest modi cation method that yielded some promising results. Thus, it portrayed effectiveness and showed that catalyst might be excluded for future BC modi cation, especially when intended for biomedical applications, and, therefore, support our hypothesis.

Declarations
The authors are unequivocally and solemnly declaring that the manuscript has not been published elsewhere and is not under consideration by any other journal. There is no con ict of interest regarding the article.

Data Availability
All data generated or analysed during the study are included in this article and are available from the corresponding author on reasonable request.